On the spectrum of climate bets, most companies pick a lane: capture a ton of CO2, store a kilowatt-hour, electrify a furnace. Recast Systems has picked a stranger one. The San Francisco company, founded in 2025, says it is building "sensing, modeling, and operational infrastructure for precise weather modification" [Recast Systems website]. Translated into plainer English: hardware in the sky, models on the ground, and a serious attempt to nudge storms before they ruin somebody's week.
Weather modification is one of those phrases that sounds either ancient or science-fictional, depending on who is talking. Cloud seeding has been practiced since the 1940s, mostly by water utilities and ski resorts looking for a few extra inches of snowpack. What appears to be different at Recast is the framing. One team member's LinkedIn headline reads simply, "Preventing severe weather events" [LinkedIn]. That is not a snowpack play. That is a hurricane, hailstorm, and wildfire-weather play, which is a much larger, much more contested category, and one where the insurance industry has been quietly desperate for new tools.
The bet
Recast describes itself as working across atmospheric science, software, hardware, and field operations [Recast Systems website]. The combination matters. Plenty of academic groups model the atmosphere. Plenty of contractors fly seeding aircraft. Very few organizations try to own the full stack from sensor to model to intervention to measurement, and that integration is probably the only way to make a defensible product in a field where the central scientific challenge is proving you actually changed anything. If you cannot measure the counterfactual storm that did not happen, you cannot sell the service twice.
The company is currently hiring a Director of Engineering through Greenhouse [Greenhouse], and a Machine Learning Researcher focused on weather modeling through Dover [Dover]. Those two roles, taken together, sketch the shape of the near-term product: instrumented field systems that generate proprietary data, feeding models that decide when and where to act. It is a hardware-plus-software business in the most literal sense.
Why it could be big
The market context is unusually favorable for someone willing to take this on. Global insured losses from natural catastrophes have run above $100 billion in each of the last several years, and reinsurers have been repricing or exiting weather-exposed regions entirely. A credible technology that reduces the severity of a hailstorm over Dallas or weakens a convective system before it organizes would have buyers lined up before the demo finished. Governments are also stirring. The UK's ARIA agency has funded climate intervention research, and several US states fund operational cloud seeding programs today.
Back of envelope: a single severe hail event in a Texas metro can run insurers $1 to $3 billion in claims. If a precision intervention costing, generously, $10 million per deployment reduced expected hail damage by even 10 percent across a season of, say, six events, the gross value created sits in the high hundreds of millions against single-digit-millions of cost. The unit economics, if the physics cooperate, are not subtle. The physics cooperating is, of course, the entire question.
The team
Public profiles point to Olivia Li, whose stated mission is preventing severe weather events [LinkedIn], and Kiran Kling [LinkedIn] among the early team. The company is recruiting from atmospheric science, ML, and field engineering backgrounds simultaneously, which is the right shopping list for the problem and a hard one to fill. The Director of Engineering posting suggests they are staffing for actual deployments rather than a pure research phase [Greenhouse].
The honest counterfactual
What skeptics will say, fairly, is that weather modification has a long history of overpromising. Decades of cloud seeding studies have produced effect sizes that are real but modest and statistically slippery, and the field has had to fight off both genuine charlatans and conspiracy theorists for so long that serious scientists tread carefully. Any company in this space inherits that baggage. The bull answer, supported by Recast's stated approach, is that the bottleneck has historically been measurement and modeling, not the intervention itself. Modern radar, satellite data, and ML weather models (the same generation that has produced GraphCast and Pangu-Weather inside the big labs) make it plausible to detect smaller signals against natural variability than was possible even five years ago. Whether Recast can convert that scientific tailwind into a paid, repeatable service is the entire company.
Regulatory exposure is the other live wire. Weather modification touches water rights, downwind liability, and a patchwork of state and international rules. A serious operational business will need a serious regulatory and legal posture from early days, which is part of why hiring a Director of Engineering who has shipped in regulated environments matters more than usual.
What to watch
The next twelve months should bring three signals worth tracking. First, any disclosed pilot, ideally with a named insurer, reinsurer, water utility, or state agency, would move the company from interesting to investable in most observers' minds. Second, a first institutional round, which the current hiring pace suggests is either recently closed or imminent. Third, any peer-reviewed or preprint publication from the team, which in this field functions as both science and marketing, because customers in insurance and government will not write checks against a website alone.
Recast Systems is attempting something genuinely ambitious in a category most climate investors have avoided because it is hard to underwrite. That is also why, if it works, there will not be ten of these companies. There will be one or two. The incumbent Recast must beat is not really another startup. It is the North American Weather Modification Council's network of established seeding contractors, who have decades of operational relationships with water districts and a deeply pragmatic, low-tech approach to the same sky. Recast's pitch, implicit in every job posting, is that the sky deserves better instruments.